coordinates closely resemble those of grunerite. (Finger 1969c); the articulation requirements in- dicate that the tetraledral chain rotation is a function of the ratio ...
THECANADIAN IAINERATOGIST Journatof the MineralogicalAssociationof Ganada Nurnber I
February 1979,
Volume 17
Canadian Mineralogist Vol. 17, pp. l-10 (1979)
X. REFINEMENT OF IHE AMPHIBOLES. - CRYSTALGHEMISTRV THE
oF rHEcRVsrALsrnucruREoF FERRoGLAUcoPHANE
MODELFOR CLINOAMPHIBOLE$ AND AN IDEAL POI.VTIEONAL F. C. HAWTHORNE llniversity ol Manitoba, Winnipeg,Manitoba R3T 2N2 Departrnento! Eartlt Sctences,
tre b compteur (1015 r6flexions observ6es), par la mdthode des moindres carr6s ir matrice complEte' jusqu'au r6sidu R = 0.032. L'ordonnance des ca' tions ressemble i celle que montre le glaucophane Gapike & Clark 1968). Laffinement du ferroglaucophane indique une faible quantit6 d'Al en M(1)' en M(2), ou en M(l) et M(2\ i la fois; les longueurs moyenn€s de liaison font penser que Al est fortement ordonn6 en M(l). On pr6sente un moddle g6om6trique id6alis6 des clinoamphiboles, semblable aux moddles de Thompson (1970) €t de Papike & Ross (1970), mais qui met en oeuvre l'extension plut6t que la rotation des chalnes de t6traEdresDans ce modEle. les coordonn6es atomiques sont trds proches de celles de la grunerite (Finger 1969c), et les exigences d'articulation montrent que la rotation des chaines de t6trabdres est fonction du rapport entre deux longueurs moyennes de liaison (celle de I'octabdre et celle du t6traedre), compte tenu des perturbations dues i la rdpulsion des cations entre eux et b leur mise en ordre. Les exigences des forces de liaison au voisinage de M(4) semblent impliquer des rotations de t6traddre du type O, car des rotations .i contribueraient ir fournir a 0(6) un exc6dent d'unit€s de valence.
ABsrRAcr Three-dimensional counter-collected single-crysleast-squares tal X-ray data and a full-matrix method were used to refine the crystal structure and cation-ordering pattern of a ferrqglaucophane [a s.s87(4\, b 11.ti32(7), c 5.315(2) A, fl 103.47(3)" ; C2l ml from Calabria, Italy. The flnal R index for 1015 observed reflections is 3,2Vo. The cation' ordering pattern in ferroglaucophane is similar to that exhibited by glaucophane (Papike & Clark 1968). The site-population refinement for ferroglaucophane indicates a small amount of Al in M(l\, or M(3), or both; mean-bond-lengthconsiderations suggest that this A1 is strongly ordered in M(l\. An ideal geometrical model is presented for the clinoamphiboles, similar to the models presented by Thompson (1970) and Papike & Ross (1970\, but involving extended rather than rotated tetrahedral chains. In this ideal structure, tle atomic coordinates closely resemble those of grunerite (Finger 1969c); the articulation requirements indicate that the tetraledral chain rotation is a function of the ratio of two mean bond lengths (octahedral m.b.l./tetrahedral m.b.l., allowing for various perturbations due to cation--cation repulsion and cation ordering. Bond-strength requirements in the neighborhood of M(4) suggest that tetrahedral rotations must be O rotations, as ,l rotations would contribute excess bond strength to 0(6).
(Traduit par la R6daction)
INTRoDUcTIoN '
SovrMerns On a affin6 la structure et la mise en ordre des cations d'un ferroglaucophane de Calabre (Italie) sur cristal unique [a 9.587(4), b 17,832(7), c 5.315(2)4, I 103.47(3)"; C2/m\, i partir de donn6es tridimensionnelles recueillies sur diffractom0-
I
Glaucophane -formulais a bluish grey amphibole with Naz(Mg,Fee+)nAlrSi"Gr(OH)r. an ideal Its parageneses are virtually confined to blueschist-facies rocks (Brnst 1968, 1973') thought to have been generated under high pressure-Iow temperature conditions (Ernst 1971, Coleman 1972). In,most glaucophanes, there is some re' placement of Al by Fe3+, and complete solid
THE CANADIAN MINERALOGIST
solution apparently exists between glaucophane and magnesioriebeckite.Until recently, natural occurrencesof amphiboles in the compositional field of ferroglaucophane were not known. However, Black (197O), Hoffmann (1970), Bocquet (1974, and Makanjuola & Howie (1972) have reported analyses of ferroglaucophane, and Hoffmann (1972) has synthesized the ferroglaucophane end-member and investigated its stability relations. In this study, the structure of a natural ferroglaucophane has been refined in order to characterize the cation ordering pattern in the structure and to provide precise interatomic distances.
firmed on two other crystals by the same method. A single crystal measuring O.12 x 0.10 x 0.09mm was used to collect the intensity data. The crystal was mounted on a Syntex P2r automatic 4-circle diffractometer operating in the 0-2A scan mode with the scan rate varying between 20 and Z4olmin depending on the peak count and the or-tz soparation, using graphite monochrcrmatized MoKa radiation (tr = 0.710694). Two standard reflections were measured every 20 reflections to check for constancy of crystal alignment; no significant change was observed during data collection. A total of 1267 reflections were measured over one asym.metric unit out to a 20 value of 60". The data were corrected for ExPERIMENTAL absorption, Lorentz! polarization and background effects. and reduced to structure factors. A reThe crystals used in this investigation were flection was considered observed if its intensity kindly supplied by Dr. Christoph Hoffmann, exceedsthat of three standard deviations based University of Gottingen. Dctails of the occuron counting statistics. This resulted in lZ57 fence, chemistry and stability are given by Hoffunique reflections of which l0l5 were conmann (1970, 1972). Single-crystarlprecession photographsexhibit monoctinic-symmetry with sidered as observed. systematicabsenceshkl, h+k=2ref l, consistent with the space groups C2/ m, C2 and Cm; in REFINEMENT lqree-lnent with previous work on glaucophanc (Papike & Clark 1968), rhe space group tZ/m Scattcring factors for neutral atoms were was assumedand found to be satisfactorv.Cell taken from Cromer & Mann (1968), with anodimensions were determined by leasrsquares malous dispersion corrections from Cromer & refinement of 15 reflections aligned autbmat- Liberman (1970). The atomic parametersof ically on a 4-circle diffractomeGr: the values glaucophane(Papike & Clark 1968) were used obtainedare given in Tabte I togetherrvith other as input to the least-squaresprogram RFINE information pertinent to data coilcction. reduc- (Finger 1969a). Quoted R indices and anisotion and refinement, These values wcre con- tropic temperature factors are of the form given in Table 1. Initial site-occupancieswere assigned by TABLE l. lltScELLAflEoUS INF0$iATI0N analogy with those of glaucophane (Papike & Ch@lcal Analyslsr Clark 1968). All Al was assignedto M(2), Unit Cell Contents* \-ray data Na and Ca were assignedto M(4), and Mg 5 1 0 2 54.63 st 7.94 a(R) (r)l s.sau and Fe were assumedto be equally distributed 't7 Ar^0- l't .02 lt lv b ( E ) .832(7' ,0...0! T.t02 0.06 and M(3)l the small a.mounts Tetrahedfal s!9 c(E) 5 . 3 1 5 ( 2 ) between M(1) Fe20X 2 . 7 6 nt IV (o) 1.33 of M(2) and M(4) not occupiedby Al, Na and rot.tllllo Fe0 16.02 TI 0,01 v(i3) 883.64 Ca were assumedto be occupied by Fe. The In0 0.08 F"3* 0 . 3 1 Sp.Gr. C2ln atomic populations of the M(1) and M(3) llg0 4.75 Fe2' I .94 z 2 sites were assignedas variable with their bulk Ca0 0.98 HN 0.01 D",r. 3.22L lla20 6.25 l.l9 chemistryconstrainedto the initial value (Finger pii'-l1 27.3 L0-3 0ctahedra l 5 . 1 3 Rad./ilono. lbll u0 0 . 0 1 1969b). Severalcycles of refinement,gradually cd 0 . r 5 No, of Fobsl 1257 incrcasing I the number of variables,resulted in Na No. of Fobsi> 3a l0l5 I convcrgenceat an R index of 3.8Vo for isotropic g F l n a lt ( o b s . ) 3.2x tcmperature factors. Temperature factors were Flnal Rr(obs.) 3.4x -l'".r.1 n'Et l/ x ;r*,1 FinalR (al1 dat.) 4.51 converted to an anisotropic form as given in lro*l Flnol Rr(al1 data) 4.4r Table l, and several cycles of least-squares -1'",,.1 ,,"fE,t;roo,1 r'/t4o.l n,*, rcfinement, gradually increasing the number ToFrrture factorfom used: *t of variables,resultedin convergenceat an R of F' L_ 3.5o/o. At this stage, the equivalent iso{ropic 'frm *catculated Hoffmn (t972)i on the brsls of 23 orygens p.f.u. tcmperaturefactor of the M(2) site was signi-
,*,"u'"]
CRYSTAL STRUCTURE OF FERROCLAUCOPHANE
real. and that either there is a small amount of Al occurring in the M(1) and M(3) sites or that the amount of octahedral Al indicated by the analysis is slightly too large. As tle scattering factors of Mg and Al are virtually identical, the negative M(2) Mg occupancy corresponds directly to a reduction in the amount of octahedral Al in M(Z). The M(L) position was modified so as to be occupied by Al and Fe only, and the distribution of Fe over the four M siteswas refined with a bulk compositional constraint, together with al'l other variables. Convergencewas obtained at R and R- indices (observed data) of 3.2/o respectively' Final atomic positions and equivalent isotropic temperature factors are given in Table 2 and anisotropic temperature factor coefficients are given in Table 3. Interatomic distances and angles, and the magnitudesand orientations of the principal axes of the thermal ellipsoids were calsulated with the program ERRORS (L. W. FORFERROGLAUCOPHANE TABLE2. ATOMIC PARATTETERS Finger, pers. comm.) and are presented in Tables 4-7. Structure factor tables rnay be z Bequ'lv. x Y obtained at nominal cost from the Depository 0 ( r ) o . r o 8 e ( 30) . 0 e 4 7 ( l ) 0 . 2 0 1 6 ( 5 )0 . 6 5 ( 4 ) of Unpublished Data, CISTI, National Research 0 ( 2 ) o . l r 7 8 ( 3 )0 . 1 7 3 0 )( 1 0 . 7 4 7 8 ( 4 )4 . 5 e ( 4 ) Council of Canada, Ottawa, Ontario KlA 0S2.
ficantly less than those of M(l) and M(3). Examination of the octahedral-sitetemperature factors in ordered amphiboles (Papike et aI. 1969, Hawthorne & Grundy 1976) indicates that they are generally equal, and this has also been confirmed for disordered and partly ordered amphiboles (Finger 1969c, Hawthorne L976, Hawthorne & Grundy 1973a). This suggestedthat the scattering power at the M(2) position was not correct. Thus the Fe distribution was refined over all four M sites with the bulk composition constrainedto be equal to that indicated by the chemical analysis.This reduced the R factor to 3.3Vo and tle octahedral equivalent isotropic temperature factors became equal. However, the Mg occupancyof the M(2) site became negative, with the M(2) Fe occupancy equal to 0.152(6); this is exactly equal to the amount of Fe3* from the chemical analysis. These factors suggest that this effect is
0(3) o(4) 0(5) o(5) o(7) T(r ) T(2) r4(l) M(2) M(3) M(4) TABLE3.
Atom /n
o . 7 0 7(77) 0 . 8 5( 5) 0 0.n2e(4) 0 . 3 6 e 5 ( 30) . 2 5 2 0 ()l o . B 0 6 4 ( 5 )o . 6 e ( 4 ) 0 . 3 5 5 0 ( 30) . r 3 0 7 ( r ) 0 . 0 8 8 4 ( 5 )0 . 7 4 ( 4 ) o . 3 3 e B ( 30). r 2 2 4 ( t ) 0 . 5 7 e 3 ( s )0 . 7 5 ( 4 ) 0 . 3 0 2 ? ()7 0 . 8 6 ( 5 ) 0.3288(4) 0 0.2824(1)0 . 0 8 7 3 2 ( 50). 2 e 2 2 ( 2 )0 . 4 5 ( 1 ) 0 . 2 e 2 6 ( 10) . 1 7 2 6 8 ( 5o). 8 o 7 e ( 2 )0 . 4 5 (r ) 0.09r76(7) 1/2 0.60(2) 0 o 0 0.58(3) 0 .r 8 16 8 ( 7 ) o o 0 0 . 6 r( 3 ) 1.?7(5) 0 0.2772(1) 1/2 COEFFICIENTST FACTOR TEI|IPERATURE ANISOTROPIC FORFERRO6LAUCOPHANE
/zz
0 ( l) 1 6 7 ( 2 5 58(7) ) 0 ( 2 ) r 8 3 ( 2 5 )5 3 ( 7 ) o(3) 276(40)6 5 ( i l ) ol4) 254(26) 46(7) 0 ( 5 ) r e 3 ( 2 6 )7 3 ( 8 ) 0(6) 223(26) 7 5 ( 8 ) o(7) . 302(4r ) 3e(e) T ( r) 1 6 e (0r) 3 4 ( 3 ) T ( 2 ) r 5 7 ( e ) 35(3) !4(r) 230(1r ) 45(3) M ( 2 ) 1 8 5 ( 1 5 )42(4) ir(3) 253(r4) 3 4 ( 4 ) M(4) 488(2e)8 3 ( 7 )
* Bts-8r,xtos
lts
E.t,
552(82) -9(l',l) 4 ? 5 ( 7 6 ) - 6 ( 1 r) 700(124) o 581(8r ) -22(11) 580(77)-15(ll)
At
/zt
r4(37) 8(20) 6 r( 3 5 ) - 5 ( r e ) e8(58) o 1 3 r ( 3 e )- 1 3 ( 2 0 ) 7e(37) 66(?0) s3(38) -63(20) 4 7 2 ( 7 7 ) l 4 ( 1I ) o 5e(6r e o 2 ( 1 2 6 ). o ) 4 3 (3 r) 3(7) 308(28) I (4) 45(r3) -2(7) 323(28) -7(4) 92(14) o 413(34) o 63(le) o 0 505(46) 62(18) 0 47e(43) 0 402(371 o 12le(87) o
DtscussloN The results of the site-population refinement are given in Table 8. The refined M(2) occupancy of 0.152(6) Feo indicates that 0.134A1 FOR DISTANCES INTEMTOMIC TABLE4. SELECTED FERROGLAUCOPHANE
lengtn(8)
length (ff)
T ( )r - o ()r r r ( r ) - o ( 5 r)
T (1) - 0 ( 6 ) I T (r ) - o ( 7 ) 1 Mean
M ()r- o ( ' r ) 2 r.10 )-0(2) 2 M ()r- o ( 3 ) 2 2 M ( 2 ) - ol () M(2)-o(2) ? M(2)-o(4) 2
Meanl0
'I
.617(r) r(2)-0(6) .520
4 4 2 2
2 () ' | . 0 3 03 .e50(2) 1.8s0(3) r ,943 2.809(3) 3.244(3) ?,549(4) 3 . 6 9 0( 4) ? o6c
2,931
*bond nultipliclty
I I'I 'I
Hean
z . o e 3 ( z ) r'r(3)-o(l ) 3)-o(3) 2 .r0 r( 2 ) r'r( 2 .t 0 6
l.lean
Mean A-o(5) A-0(6) A-o(7) A-0(7) lileanl2
1.625(3) r(2)-0(2) L 6 1 5 ( 3 ) T ( 2 -) o ( 4 ) '1r. 5 2 3 ( 3 ) T ( 2 ) - o ( 5 )
Hean
H(4)-o(2) M(4)-o(4) 11(4)-o(5) M ( 4-)o ( 6 ) l'lean 6 Hean 8
M ( l) - r 1 ()r M(',r )-H(2) M ( )r - M ( 3 ) r,r( r)-H(4) N(2) -r't(3) lr(2)-r,r(4 )
4
(3) r .631 r . 5 e 63( ) 'r (3) .65r I 6E6/?\ .ry
I .633 2.t38(2) 2 . o e o ( 4) 2.122 2 . 4 0 3 3( ) 2.33?l3l 2 . 8 2 8 (3 ) 2 . 4 5 7( 3 1 2,397 2.505 3.272(2) 3 . 1 0 a ( )r 3.121(r) 3 . 3 0 7 (2 ) 3 . 2 4 0 (r )
3.r57(r)
THE CANADIAN MINERALOGIST
TABLE5. *
(8) FOR POLYHEDRAL EOGELENGTHS FERROGLAUCOPHANT Distance
T( l ) tetrahedron
2 . 6 4 (63) 2 . 6 6 (53) 2 . 6 5 7( 4 ) 2 .6 5r ( 4 ) 2 . 6 3 (r 3 ) 2,622(3) ?..646
0(1)-0(5) 0 (l ) - 0 ( 6 )
o (r)- o ( 7 )
0(5)-0(6) 0(5)-0(7) 0(6)-0(7) Mea n
l,l( l ) octahedron
o(r:)-o(2:) z 0 (r : ) - o ( 2 : ) 2 o(r:)-o(3:) 2 0 ( 1 " ) - 0 ( 3 " )2 o(2)-o(2) 1 0(2)-0(3) 2 0(3)-0(3) r }lean
2.64s(3) 3.204(3\ 2.u4(4') 3.16e(4) 3.045(5) 3.0e1(2) 2.70e(7\ 2.973
fl(?) octahedron
o(r)-o(r), r 0(r:)-0(2:) 2 0('r")-o(2") 2 0(r).-0(4), 2 o(2:)-o(4:) 2 0(2")-0(4") 2 0(4)-0(4) | Mean
2.52r(5) 2.64s(3) 2.80s(3) 2.742(4) 2.811(3) 2.6es(3) 2.845(5) 2.740
Atoms
r
TABLE7.
Distance
T(2) tetrahedron
0(2)-0(4) 0(2)-r(5) 0(2)-0(6) o(4)-o(5) o(4)-o(6) 0(5)-c(6)
2.748(3) 2.664(3) 2.652(3) 2 . 6 5 33() (4 ) 2. 5 9 1 2 . 6 8 04() 2.665
Mean
M(3) octahedron
o ( r : ) - o ( r : )z 0 ( r : ) - 0 ( r : )2 o ( r : ) - o ( 3 : )4 0(r")-o(3'4 )
pMc Auurrr
0 ( l) 0(2) 0(3)
z.6zt(51 3.378(5) 2.u4(4\ 3.12e(4) 0(4)
lilean
2.991
14(4)polyhedron 0(2)-0(2); r 0(2:)-0(4:) 2 o(2:)-o(4:) 2 0(2:)-0(5:) 2 o(4I)-0(5:) 2 0(4:)-0(6:) 2 o(5:)-o(6:) 2 o(5")-o(6") 2 o(6)-o(6) I l,lean
3.04s(5) 2.81r(3) 3.263(3) 3.602(4\ 3.3se(4) 2.ser(4\ 3.04e(3) 2.651(4) 3.36e(s) 3,067
Mul tip l lcl ty
0(5) 0(6)
0 (7 )
r(l) 1(2\
TABLE 6. SELECTED 1NTEM1OMIC AI'IGLES FORFERROGLAUCOPHANE M(l) .
*
A n s t e( o )
A n g t e( o )
T(1) tetratEdron
T(2) tetrahedrcn
0(r)-T(r)-0(5) I 10s.5(t) 0(1)-T(r)-0(6) r il0.3(r) 0(r)-T(r)-0(7) r il0.0(2) 0(5)-T(1)-0(6) r r0e.e(1) 0 ( 5 ) - T ()1- 0 ( 7 ) r r0e.0(2) 0(6)-T(r)-0(7) r r08.r(2) llean 109.5 l4(l) octahedron
0(2)-T(2)-0(4) r 116.8(l) 0(2)-T(2)-0(5) I 108.5(r) 0(2)-r(2)-0(6) r 107.6(r) 0(4)-T(2)-0(5) 1 10e.6(l) 0(4)-T(2)-0(6) r r 0 5 . 7 ()r 0(5)-T(2)-0(6) r t08.3(r) tlean t09.4 !,t(3)octahedrcn
o t t l ) - i a ( r) - o ( a lt z 7s.4(t) 0(r;)-rf(r)-0(2:) 2 se.6(11 0(r;)-t4(r)-0(3;)2 84.8(1) e7.4(1) Q ( r l ) : ! r ( 1 ) : Q ( : - ' )? 0(2)-il(l)-0(2) r e2.e(l) ofzi+titi-oiri z' l ga.iiij 0(3)-r{(l)-0(3) 79.211\ ilean 9o'o fi(2) ocrahedron
o l r u1 - r , 1 a 1 - o 11r l 2 0(1:)-H(3)-0(t:) 2 0 ( r : ) - r 1 ( 3 ) - 0 ( 3 : )a 0 ( r " ) - i r ( 3 ) - 0 ( 3 " )4
80.4(1) 0(r)-M(2)-0(1)i r 0(r;)-fi(2)-0(2;) 2 83.4(l) 0(r:)-r'{(2)-0(2') 2 8 e . 5 ( r) 0(rJ+r(2)-0(4)i 2 8e.8(l) 0 ( 2 ; ) - 1 4 ( 2 ) - 0 ( 4 - ' )2 es.4(l) e o . s ( r) 9(?:).-.ll?)-g(1") ? 0(4)-r(2)-0(4) I 1 0 0 . 5 0) llean 89'e o(7\-o(7)-0(7167.2(21
*a
o.?s3
7 5 . 6 ( r) 104.4(t) 84.5(t) e5.5fl)
MAGNITUDE A!'IDORIENTATION OF PRINCIPAL AXEs THERMAL ELLIPSOIDS FORFERROGLAUCOPHANE
t4(z) M (3 ) i,r(4)
deviation
Angle to a-axis
Angle to b-axis
Angle to c-axis
o . o 7 e ( 7 ) ( 84) e ( 1 7 ) ( o )8 4 ( 1 8 ) ( o ) 5 5 ( et )( o ) o . o s (l 6 ) 60(24) 4 7( 2 6 ) 12.e(24) 0.ror(6) r25(le) 44(26) 5 e ( 2)r 0.076(6) e3(20) 85(18) 12(22) 0.088(6) r48(48) 122(48) 78(22) e2(18) 0 . 0 e 4 ( 6 ) 5 8 ( 4 8 ) 147(47) 0.0e7(8) e4(27) 90 1o(27) 0.102(8) e0 90 0 o . r i r( 8 ) 4(27) 90 r00(27) 0.082(7) 67(33) 23(29) 're4(88) 0 . 0 8 5 ( 6 ) 7 1( 3 e ) e 7 ( 8 2) 7r (43) 0.110(5) 31(r0) lr2(r0) 82(1r) ',r 'r2r(6) 0.076(7) r3(r6) 34(ro) 0.0e2(6) 150(ls) e4(11) 106(r6) 0.il8(5) r08(e) 3l(6) 6 1( 6 ) o . o 7(r7 ) 25(5) e 7 (r 0 ) 66(6) 0 . 0 e 7 ( 6 ) r 5 5 ( r)1 7 0 ( 1)1 e3(rr) o . i l e ( 5 ) r ' r 4 ( r)' r r 4 8 ( 8 ) 65(5) 0 . 0 7 9r() 90 0 90 0. l07(8) 11e(22) 90 r38(22) 0 . 1 2 2 ( 7 ) 1 5 1 ( 2 2 ) 90 48(22) 0 . 0 6 4 ( 3 ) e r( 5 ) e 5 ( r3 ) l3(7) 0.074(3) 8e(e) r75(r3) e5(r3) 0.087(2) 2(7) 8e(e) r 0 2 ( )5 'r4(r 0.066(3) sl(8) 86(15) r) A2l1^\ 0.074(3) |r2(ro) r58(ro) o.085(2) 2 2 ( 1 0 ) r r ( r 0 ) l 0 r( 6 ) 'r02(4) 0.073(3) 90 2(4) n on 0 . 0 8 (s3 ) 90 o.rol(2) 12(4J 90 e2(4) 0.082(3) 88(16) 90 r6(r6) 0.083(3) 90 0 90 0 . o e(]3 ) 2(16) 90 r06(r6) n 0.074(4) 90 90 0.080(3) e0(5) 90 r6 7 ( 5 ) 0 .r 0 7 ( 3 ) 0(5) 90 l03(5) o.106(5) r30(4) 26(4) 90 0.il5(s) 90 0 90 0.r54(4) 40(4) 90 64(4)
this problem. However, the ionic radii of Al'* (0.53A) and Mg:+ (0.724) differ considerably; H(4)mlyhedron thus, the mean bond lengths should give some o(2)-H(4)-o(2), r 7g.6( indication of the site occupancies. Using the 0(2:)-t{(4}-0(4:) , ,.ttli o t 4 ) + i a l - o ( r j lz a z . r ( r ) curves of Hawthorne (1978) relating mean 0 ( 2 : ) - H ( 4 ) - 0 ( 5 j )2 86.6(l) bondJength to constituent cation and anion 0(4:)-M(4)-0(5") 2 80.6(l) 0(4:)-r1(4)-0(6:) 2 5 5 . s ( 1) radii, the following calculated mean bond 0(5:)-it(4)-0(5:) 2 70.0(l) 0(5')-M(4)-0(5") 2 5e.7(1) lengths are obtained for the two extreme distri0(6)-14(4)-0(6) l 8 6 . 5 ( r) bution models: ilean 75.6 nean
T(r)-o(5)-T(z) r(l)-0(b)-T(z)
e0'0
134.5(2) 143.s(z)
ltll:3tll:l[] lii:llii 0(s)-0(7)-0(6) r 7 0 . 1( 2 )
+4 - (e00)rsoo. 'tlultrpl lo(r)-o(r)-o(z)] r"ity per formula unit must occupy M(l) or M(3) or both. As the scattering factors of Mg and Al are very similar, there is no direct information available from the refinement concernins
M (I) : 0.585Fe'z+ *0.348Mg*0.06741 M(3) -0.795Fe'z+*0.205Ms o < M (l)4> -ro= 2.098A, < M(L)-O >"b": 2.1.06A *r" :2.113A, "b" =2.1224, M(1) :9.567Pu'z+*0.413Mg M (3) : 0.795Fe'+*0.071Mc + 0.134A1
*r": 2.110i,.b": 2.1961i < M(3)-O>*r":2.0894, ( M(3)>"6": 2.1224
CRYSTAL STRUCTURE
TABLE 8.
OF FERROGLAUCOPHANE
IN FERROGLAUCOPHANE SITE-OCCUPANCIES
Fromsite-population refinement: l l ( 1 ) 0 . 5 8 5 ( 6 ) F e+o 0 . 4 l 5 M s + o.844Al M(2) 0.156(6)Feo o o.205Ms N(3) 0.795(9)Fe+ M(4) 0.B60N+ a 0 . 0 7 5 c a+ 0 . 0 6 5 M 9 Fromexaminationof meanbond lengths: g 0.067A'l M ( ] ) 0 . 5 8 5 ( 6 ) F e 2++ 0 . 3 4 8 M + ?+ H ( 2 ) 0 . 1 5 6 ( 6 ) F e "+ 0 . 8 4 4 4 1 M ( 3 ) 0 . 7 e 5 ( 9 ) F e ' '+ 0 . 2 0 5 1 ' l s t4(4) 0.B60lla+ 0.075ca+ 0.0651'19
The root-mean-square deviations for the trro extreme distributions [all Al in M(l) or M(3)J are 0.009A and 0.024A, respectively, suggesting that the octahedral Al not residing in M(2) is strongly ordered in the M(1) position. The refined M(2) occupancyof 0.152(6)Fe' is equal to the amount of Fe"* present as indicated by the cell contents (Table 1); this factor, together with the observed mean-bondlength, suggeststhat all of the iron in the M(2) bond length and site is Fe3+.T:he 1M(2)-O) constituent M(2) cation radius obtained for ferroglaucotrrhanein this study (
